Upload app.py

app.py

@@ -74,19 +74,19 @@ args = parser.parse_args()
 
 args.device = "cuda"
 
-base_path = 'feishen29/IMAGDressing-v1'
+# base_path = 'feishen29/IMAGDressing-v1'
 
 
-vae = AutoencoderKL.from_pretrained('./ckpt/sd-vae-ft-mse/').to(dtype=torch.float16, device=args.device)
-tokenizer = CLIPTokenizer.from_pretrained("./ckpt/tokenizer")
-text_encoder = CLIPTextModel.from_pretrained("./ckpt/text_encoder").to(dtype=torch.float16, device=args.device)
-image_encoder = CLIPVisionModelWithProjection.from_pretrained('./ckpt/image_encoder/').to(dtype=torch.float16, device=args.device)
-unet = UNet2DConditionModel.from_pretrained("./ckpt/unet").to(dtype=torch.float16,device=args.device)
+vae = AutoencoderKL.from_pretrained('stabilityai/sd-vae-ft-mse').to(dtype=torch.float16, device=args.device)
+tokenizer = CLIPTokenizer.from_pretrained("SG161222/Realistic_Vision_V4.0_noVAE", subfolder="tokenizer")
+text_encoder = CLIPTextModel.from_pretrained("SG161222/Realistic_Vision_V4.0_noVAE", subfolder="text_encoder").to(dtype=torch.float16, device=args.device)
+image_encoder = CLIPVisionModelWithProjection.from_pretrained("h94/IP-Adapter", subfolder="models/image_encoder").to(dtype=torch.float16, device=args.device)
+unet = UNet2DConditionModel.from_pretrained("SG161222/Realistic_Vision_V4.0_noVAE", subfolder="unet").to(dtype=torch.float16,device=args.device)
 
 # image_face_fusion = pipeline('face_fusion_torch', model='damo/cv_unet_face_fusion_torch', model_revision='v1.0.3')
 
 #face_model
-app = FaceAnalysis(model_path='./ckpt/buffalo_l.zip', providers=[('CUDAExecutionProvider', {"device_id": args.device})]) ##使用GPU:0, 默认使用buffalo_l就可以了
+app = FaceAnalysis(model_path="buffalo_l", providers=[('CUDAExecutionProvider', {"device_id": args.device})]) ##使用GPU:0, 默认使用buffalo_l就可以了
 app.prepare(ctx_id=0, det_size=(640, 640))
 
 # def ref proj weight
@@ -126,7 +126,7 @@ adapter_modules = torch.nn.ModuleList(unet.attn_processors.values())
 adapter_modules = adapter_modules.to(dtype=torch.float16, device=args.device)
 del st
 
-ref_unet = UNet2DConditionModel.from_pretrained("./ckpt/unet").to(
+ref_unet = UNet2DConditionModel.from_pretrained("SG161222/Realistic_Vision_V4.0_noVAE", subfolder="unet").to(
     dtype=torch.float16,
     device=args.device)
 ref_unet.set_attn_processor(
@@ -167,7 +167,7 @@ noise_scheduler = DDIMScheduler(
 # noise_scheduler = UniPCMultistepScheduler.from_config(args.pretrained_model_name_or_path, subfolder="scheduler")
 
 control_net_openpose = ControlNetModel.from_pretrained(
-    "./ckpt/control_v11p_sd15_openpose",
+    "lllyasviel/control_v11p_sd15_openpose",
     torch_dtype=torch.float16).to(device=args.device)
 # pipe = PipIpaControlNet(unet=unet, reference_unet=ref_unet, vae=vae, tokenizer=tokenizer,
 #                         text_encoder=text_encoder, image_encoder=image_encoder,
@@ -183,7 +183,7 @@ img_transform = transforms.Compose([
     transforms.Normalize([0.5], [0.5]),
 ])
 
-openpose_model = OpenposeDetector.from_pretrained("./ckpt/ControlNet").to(args.device)
+openpose_model = OpenposeDetector.from_pretrained('lllyasviel/ControlNet').to(args.device)
 
 def resize_img(input_image, max_side=640, min_side=512, size=None, 
                pad_to_max_side=False, mode=Image.BILINEAR, base_pixel_number=64):